Hi, I'm Athish M
Full-Stack + AI/ML Developer
2nd-year AIML student. Full-stack is my foundation, AI/ML is where I specialize. I treat side projects like production software: deployed, documented, and actually usable.
Selected Work
Click a card to view architectural details
context-gc
Deterministic context-compaction library for AI agents. Prunes obsolete and superseded steps from execution traces via graph-based analysis.
AuraVault
AI-powered wealth management and financial dashboard utilizing data visualization to analyze portfolio allocation and deliver personalized asset insights.
SafeBuildXR
AI-powered circuit tutor engineered for virtual reality environments. Registered for the XRCC Creator Conference.
WayFinder
A dynamic route planning and pathfinding navigation web application designed to calculate optimal paths and model map graph traversals.
Expense Tracker
Dynamic full-stack web application built to track financial data seamlessly, featuring a custom Flask backend and a clean vanilla JavaScript interface.
Technologies & Tools
Hover or tap a skill to highlight associated projects
Frontend
Backend
AI/ML
Database
Infra/Tooling
Specialized & Core
How I Build
I follow a four-step loop centered on delivering clean, usable code early, and gathering real metrics to refine it.
Scope
Understand the problem, constraints, and intended users. Define the smallest version that is actually useful.
Ship
Build the smallest useful version and deploy early. Real URL, real users, no 'coming soon' pages.
Instrument
Add the data, AI, logging, or feedback layer needed to monitor behavior and performance.
Iterate
Fix what breaks, optimize bottlenecks, prune what doesn't matter, and repeat.
Currently Focused On
Data Structures & Algorithms
Deepening Data Structures & Algorithms (DSA) to solve complex graph traversal and data compaction problems.
System Design Fundamentals
Sharpening system design fundamentals for full-stack + AI products, with a focus on latency, security, and edge deployments.
About & Philosophy
Behind the commits
I study Artificial Intelligence and Machine Learning formally at college, but I spend my nights diving deep into core software systems. My philosophy is that algorithms are only as good as the infrastructure they run on. I don't build isolated Jupyter notebooks or experimental scripts — I package logic models, graph routing, and financial visualization boards into responsive, documented codebases.
Treating side projects like production software means I focus on the details: deterministic algorithms that don't leak resources, strict validation logic on API bounds, and keyboard-accessible UI systems. I am constantly working to verify that when someone clicks a URL on my portfolio, it loads instantly, handles errors gracefully, and delivers a robust UX.
When I am not in front of my editor, I study high-growth tech startup architectures, experiment with edge deployments, and research compiler targets like WebAssembly to bring systems-level efficiency to the web browser.